[Home]DouglasReay/AmiCog

ec2-35-173-57-84.compute-1.amazonaws.com | ToothyWiki | DouglasReay | RecentChanges | Login | Webcomic

I have take the ideas discussed here and put them in the form of an essay:

 [http://www.chiark.greenend.org.uk/~douglasr/amicog/]




From "Amicus" (friend) and "Cogito" (to think).

thus: someone you think is your friend --AnonymousHumorist??

An Amicog (Noun): A collaborative knowledge structure formed by the interactions of a group of people.  Recordings of who believes what, and how much whose judgement in which areas is respected by whom.  The technology required to add to this, organise it and present customised views of it to participants.

To Amicog (verb): To group think.  To interact with an Amicog that you are part of.


A simple pre-cursor to a full amicog might be a Wiki combined with an Advogato?-like TrustWeb? (but more subjective), that highlighted the contributions of the users rated most highly by your TrustMetric in that area (possibly entirely hiding contributions from users who you ended up rating negatively).

That's a new abbreviation on me - can't say I like it any better than the others.  Adhocracy contains more connotations and sounds just as silly :)
On one level - how is this any different from today's group identity?  People think as part of a sports team, or on behalf of a club.
It is an added level of abstraction.  3rd order /MediatedTrust instead of 2nd or 1st order. --DR
On another - the machine assisted trust network used to increase the size of such a network, and to fuzz the boundaries brings on some nasty issues.  Not everyone wants to be a politician.  Further - any automated system can be gamed.  And if the system allows for backlash (so when people figure out they have been gamed, they hit back) then that system can be gamed, and so forth.  I'll leave aside the clumsy transition period (which isn't addressed at all) and assume that kids grow up used to interacting with the system's cues as naturally as people read body language when deciding who to trust.
I agree that all rule based systems are gamable to some extent. (eg housing benefits, immigration, asylum).  I disagree that just because a system is gamable, it will necessarily display unstable equilibrium under that influence. --DR
I'm having trouble trying to focus on my precise objection to the system.  All of the "people won't accept it" can be handwaved away, there'll be ways to bring about such a system.  I can also see that it is in many ways beneficial - I just gut feeling do not like the idea of actually participating.
So you see as potential negatives of the system that you believe it would become politicised?  Out of interest, do you believe that this Wiki is politicised?  How about usenet newsgroups (pick your own examples)? --DR
Maybe an example is best.  Guild systems in competitive MMORPGs / MUD?s.  People trust their guild members to a greater degree than non-members, guilds have trusts between each other and blacklist each other etc.  The usual evolution is that a single guild becomes more powerful than most others.  Everyone flocks to join - wanting the power, it eases its rules to let more people in.  For a time it dominates.  Eventually the leadership loses its grip, the guild fractures into two smaller pieces.  A few of the guilds opposing it team up and take in members that were lost in the split - and gains dominance itself.  Repeat until game has too few players to sustain.  --Vitenka
I'm not sure that is the right parallel.  I can see two seperate amicogs getting built in the same area (say the creation of the world) around fundamentally incompatible starting assumptions (evolution vs creationism).  I can see seperate amicogs in related areas (say Anime and Cartoons) developing overlap.  I can't see when you'd the sort of competition for members you are talking about, because wouldn't it imply that someone was changing what they say they believe, in order to fit into the new group? --DR
Being a memeber of a group confers advantages.  Thus people will game the system - and guilds have shown us that people will game the system en-masse.  Indeed, enough people joining purely for the benefit of having joined rather than the presupposed intent of the guild will change the aims of the guild.  Then all the people who already joined (proibably now outnumbered) are annoyed, and new people joining won't even know about the original aim of the guild.  And this example is from natural (what you're calling second order) trust - automate it and things can only get worse.  --Vitenka
Can you give a concrete example?  See /ToothyCog
Every guild in every multiplayer game ever, for starters.  --Vitenka
Apologies.  I meant "Can you give a concrete example of how this dynamic would apply to an AmiCog??". --DR

(Belatedly joining the discussion...)  DouglasReay said above "I can't see when you'd the sort of competition for members you are talking about, because wouldn't it imply that someone was changing what they say they believe, in order to fit into the new group?"  I'd say the competition would not be for people's opinions but for their time, their regular visits and thus their loyalty. There are many many online communities which I would like to participate in more regularly than I do, based in a variety of media such as NewsGroups?, forums, discussion boards, mailing lists, wikis, and IRC channels. I can think of at least two in each of those categories which I have participated in to some extent in the past, and for which the reason why I no longer participate is nothing to do with changes in beliefs but shortage of time. In most cases I would actively value /AmiCog style opinion matching for recommendations, whether the things recommended be Renai games, essays on theology, interesting WikiPedia pages, or whatever. However, I don't have time currently to interact with all these in their current technological state, let alone if I had to manually add ratings. So /AmiCogs? would be competiting for my time and loyalty, the way several online communities and sites already do. (And other sites: eg Amazon begs me to review my purchases, etc). --AlexChurchill
Welcome to the discussion.  This is the eyeball theory of the [attention economy].  Let's see how it would apply to an AmiCog?.  Suppose you were interested in the ressurrection.  You believed not just in the spiritual but also in the bodily ressurrection of Jesus Christ, and had written a think piece supporting that opinion.  And now you publish it (both your backing of the opinion and the text).  Publishing to multiple places tends to scale better than linearly in the time required, because of automation and gating, so let's assume you send it everywhere.  You put it on your website, your live journal, ICQ it to a friend, send it to your local church email list, add it to a usenet discussion, place it off the ReligiousMatters? wiki page and amicog it, all signed appropriately for your identity in each forum (ok, the publishing client that can do all that hasn't been written yet, but think 30 years ahead).  What next?
Next you recieve feedback.  People send you emails, post response usenet articles, respond to the wiki article and add their opinions to the amicog and rate yours.  You now have to decide whose feedback (where) you spend your time reading.  Who will you respond to, who will you rate (explicitly mention you agree or disagree with their particular points), who will you make references to in your follow up article?
When the choice is between spending time on two different amicogs, both of which have responses to (ratings of) your original assertion of opinion, I think there may be a certain flexibility.  Because the data of an AmiCog? doesn't have to be centralised, and because each person's view of it can be personalised (according to the scores allocated to each source by their personalised trust metric), you can have slightly fuzzy edges of who (and what articles) are part of which AmiCog?.  So you can happily give feedback to particular individuals who have made comments on your original think piece that you find worthy of response, irrespective of which AmiCog? they made them in. --DR



People are going ahead with the socially-influenced search. [This article] on SearchEngines? reports that in 2005/6,
Today a number of startups are attempting to infuse human preferences into computer algorithms to pinpoint relevant information better. For instance, somebody searching for a nearby bicycle store may benefit from knowing which shop's site has been most frequently bookmarked by people in the neighborhood, or described in the most favorable terms. "It gets interesting when human knowledge begins to intermingle with algorithmic search," says David Hayden, CEO of startup Jeteye Technologies Inc.

And more. [This article] describes situations like this:
A stranger on the street will ask you to loan him $20, and you'll actually seriously consider his request. Why? Because you can see his name and address. More importantly, you can ping his whole social network and see how many of those who trust him are people you trust, or are trusted by people who are trusted by people you trust...
a world where everyone has an online diary or they're no one. Not much I agree with Sartre? on, but Hell really is other people's LiveJournals?. --ChiarkPerson
Good Quote?, but everybody already has a FootPrint? - making it electronically searchable rather than having to ask their freinds, their friends-freinds etc. seems inevitable.  And makes it possible to automate scams.  Hmmm.  --Vitenka
Lies. How many of your old schoolmates can you Google? Not everybody is on the web. Neither should they have to be. - ChiarkPerson
I think Vitenka meant everyone has an impact on the world - even if an individual is not googleable, there will be a chain of people linking you to that individual. I'm unconvinced it would be the proverbial 6 links long, though - SunKitten
Aye - the information exists.  In a few years time it will almost all be captured electronically, and in a few years after that - searchable.  (Insert your own timetable for 'few years')  Whether this is desirable or not doesn't really matter.  It doesn't seem realistic to hope that it all goes away.  Better to start planning how to make it not suck.  --Vitenka
See what it is is, these articles are written by internet journalists who are excited about the internet. And they do all their research on the internet, so all they see is other peopel who are on the internet, and they're on the internet because they're excited about the internet. It's a horrendous self-inflating vicious circle of cyber-egos. They only good thing is that while they're all talking to themselves they leave the real world, for whom the internet is a neat way to waste time and do some shopping while at work, alone.
The feedback cycle and echo-chamber comments are valid.  But you're thining in far too limited a way.  Information does not imply that the internet is in any way involved.  --Vitenka

[This article from Wired] and [this post from Scooble] show that people are starting to realise why they don't want social networking sites to own the user's data.  The data storage and the data presentation are seperable layers, and ought to be seperated.
Ditto [the conflab over Microsoft using contacts export as leverage].  See also [Tim Berners-Lee on the seperation of links and metadata from the data]

It also suggests googling for "[Reputation Economy]".
SeeAlso http://www.last.fm, Pandora, etc... --AC

[The Economist on social networks magnifying existing networks]


CategoryFuture
See also: /TheFuture /DistributedComputing /LivingApplications /KnowledgeStructures /MediatedTrust /SocialConsequences /AmiCog /ToothyCog

ec2-35-173-57-84.compute-1.amazonaws.com | ToothyWiki | DouglasReay | RecentChanges | Login | Webcomic
Edit this page | View other revisions | Recently used referrers
Last edited May 24, 2011 6:59 am (viewing revision 31, which is the newest) (diff)
Search: